1. CSCD 506 Research Methods for Computer Science
Fall 2018
Lecture 1
Course Overview 1 1

2. Topics
Goals of the Course
Activities
Your role
My role
Grades 2

3. Course Trivia Website http://penguin.ewu.edu/cscd506
Not using Canvas for this course
Meeting Times and Days:
2:00-4:30 pm Mon, Wed
Instructor
Office hours: Tues, Thurs 1 – 2 pm
Anytime I am in my office

4. Goals of this Course
Overall, to learn how to do research in Computer Science
More specifically ….
To learn how to conduct literature reviews including critically reading papers, assessing their results and summarizing the results in a written form
To learn how to pose a researchable question
What constitutes a chunk of research at the Masters level? 4

5. Goals of this Course continued
To learn the difference between a project and a Thesis
To learn how to write a publishable paper within the field of Computer Science. And, by extension, how to write a thesis or project report
To learn how to use statistics to support your work 5

6. Activities to Support the Goals
Research Activities
Conduct a Literature Review
Define a Research Question
Write a Research Proposal
Previous steps feed into the Proposal
Analyze some data sets and learn to use statistical software
If there is time and interest
Presentation of your Proposal 6

7. Conduct a Literature Review
Reading papers relevant to a topic you are researching
What is a Literature Review?
And, why do you do one?
A Literature Review
Surveys and summarizes scholarly research on your topic. It should critically evaluate and explain why the chosen articles are significant in relation to your research topic
You use the results to better shape your research
Knowing what others have done allows you to find the “gaps” in the knowledge where you can contribute 7

8. Come up with a Research Question
What constitutes a “doable” chunk of research?
No hard and fast rules for this
Guided by time-frame and ambition of researcher
If you have 3 – 5 years to complete your project, you do more, ask harder questions
If you have 1 – 2 years, you ask a smaller question
Ultimately, your advisor guides you on extent of your problem, but it should ideally be your idea 8

9. Masters vs. PhD Masters degree PhD Thesis: Learn to conduct research
Experiment, but doesn’t have to be original, could reproduce someone else’s work or extend it a small amount
Project: Produce a significant piece of software or system
PhD
Needs to be original. Assume already know how to do research. Add to body of knowledge in a significant way. Exactly how and what depends on your committee. 9

10. Thesis Examples Sampling of Computer Science Thesis's
University of Nebraska Lincoln
1. An Unmanned Aerial System for Prescribed Fires by Evan M. Beachly 166 pages
2. Significant Permission Identification for Android Malware Detection by Lichao Sun 61 pages
3. Bandwidth Estimation for Virtual Networks by Ertong Zhang 138 pgs
Cal State University Northridge
1. A study and comparison of NoSQL databases Choi by Christopher Jay 74 pages
2. ECG and wearable computing for drowsiness detection by Gupta, Nipun 71 pages

11. Dissertation Examples
A sampling of Dissertations
1.Inference in Networking Systems with Designed Measurements by Chang Liu 174 pages
2.Knowledge Representation and Reasoning
with Deep Neural Networks by
Arvind Ramanathan Neelakantan 121 pages
3. Fundamental Limits of Covert Communication by Boulat A. Bash 166 pages
Plus many others …. Umass Amherst

12. Research in CS Computer Science Research is Diverse
Designing and building new computer systems,
Proving mathematical theorems,
Writing computer software,
Measuring the performance of a computer system, using analytical tools to assess a design,
Or studying the errors programmers make as they build a large software system
Can also study users of systems as in Human Computer Interaction or Security Usability 12

13. Research in CS
Researcher chooses activities appropriate to answer questions that arise in research investigation,
New questions arise as an investigation proceeds,
Research activities vary from project to project and over time in a single project
A researcher should be prepared to use a variety of approaches and tools 13

14. Define a Research Question
The hardest part of the Masters degree!!!
What am I going to do?
We will explore steps to help you define your question
A few of the questions you need to ask are on the next slide 14

15. Define a Research Question
What PROBLEM you are trying to solve?
Or what is research QUESTION you are trying to answer?
Why is this problem/question worth solving/asking? Who would care?
How have other people tried to solve or answer it?
What is your NEW approach to solving or answering this problem?
Or What improvement are you making on an existing solution? 15

16. Research Question
Defining the right question helps you craft a solution
Doable
Of sufficient difficulty to constitute research or at least a different approach
Interested in answering the question or solving the problem
Feeds into the Research Proposal

17. Write a Research Proposal
Steps to Accomplish This
Review other's work, become familiar with your chosen “field”
Reading papers from journals, conferences, experts in the area
Define, or refine your question or problem you want to solve
Create your solution
Write a proposal that presents your idea for solving the research question 17

18. Analyze Data sets with Statistical Software
For research problems, need to know how to analyze data
Computer science research doesn’t always require statistical analysis
Depends on the problem
If collect data as part of an experiment, likrly need to analyze data
Will look at courses available at EWU for further study on statistics 18

19. Present Proposal Will have chance to do
a presentation of your research topic
Presentation skills are invaluable for everything you will do in the future
Both non-academic job activities and academic
If you can’t present your ideas, in a way that is understandable, no-one will care about them 19

20. Your Role Actively participate in class
Contribute links and any other interesting material you think will help yourself or other students
I am one person
You are many ... will likely discover resources that I have not encountered
Try hard to choose a topic you can use for your research 20

21. My Role Guide to the literature on Computer Science research
Show you helpful articles, a few books and other resources
There are many resources at other schools, from other professors and within ACM/IEEE literature
I have not written any of it …
What qualifies me to guide you?
I have done research in pursuit of both my Masters and PhD
Somewhat qualifies me as a guide
Will use other’s experience as well 21

22. Acing the Class Put in an honest effort
Do the assignments, contribute to the class
Might not finalize masters topic by the end of the quarter
But, you should try!!!
This is something you need to do anyway, and you get credit for it
Should use this opportunity to practice writing skills before it really counts 22

23. Look more at CS Research
Next time,
Look more at CS Research
End 23